Avi Loeb has a post on the WiA blog that echoes his arXiv posting about “how to collect matches that will catch fire.” It’s a metaphor for identifying talent among young, unproven professionals.
I know that blog post and paper are getting a lot of justified criticism, but it got me thinking about similar metaphors for identifying talent. It reminds me of an old baseball yarn about how to scout talent (I think I heard this first from Nate McCrady; not sure).
You’re a talent scout watching two young ballplayers. Their stats are almost identical, and in particular they run from the batter’s box to first in about the same amount of time after a hit.
An astronomy student taking a physics exam as an admissions committee looks on.
One, though, has beautiful form, a perfect stride. Really looks like a pro. The other has terrible form: arms flapping all around, not adjusting his stride to stretch for the bag — he looks like an amateur.
What does your report say? Who do you recommend your manager try to sign?
The gut-based answer is you want the great runner — he reminds you of a pro, and your gut likes to stereotype by preferring things that look like success.
The, brainy, numbers-based answer is that it doesn’t matter; despite appearances, they’re actually equivalent. Form is irrelevant — go with function (speed). You should be blind to appearances when what you want is speed. This feels nice and objective and fair. Who cares what he looks like when he runs?
But the right answer is that you go with the sloppy runner. The good runner has clearly been taught how to run, but he’s still no better than the sloppy runner, who clearly doesn’t know what he’s doing and is going on pure talent. The first player has little remaining room for improvement, but the second player—teach him to run, and imagine how fast he’ll be then.
In other words, if two people have identical stats, but one has better preparation, then the better long term bet is the one with less preparation.
As Avi says:
a young researcher who did not benefit from the privilege of being nurtured by top quality mentors or had to transition from a different culture or an [lower] socio-economic status, should be given more slack. This is common sense.
But a purely metric-based appraisal of young researchers, also feels like “common sense”. And preferring students from the “best” schools also feels like “common sense” (a better education!). Common sense is not self-consistent, so it’s not really “sense” (but it is common!).
Yes, that’s exacly what I was thinking, too.
Great post Jason. I had another thought while I was reading it, through the lens of all the diversity discussion that has been percolating through the field recently. “The gut-based answer is you want the great runner — he reminds you of a pro, and your gut likes things that look like success.” I wonder if the same analogy applies to two people interviewing for a position with the same ‘stats’, but one is a man and one is a women (or choose another majority/minority contrast). If what ‘looks like success’ to the panel, even unconsciously, is the person from the majority, then it is a good way of demonstrating/explaining the bias.